dnak
04-25-2007, 10:23 AM
Hi, just wondering...
We have just been assigned a big job including rigging and building + lipsynch character and putting this into live footage, handheld camera. The spot will be approx 3 minutes long including separate graphics and we will also be in charge of the shoot on location. What we want to achieve is something similar to www.ericlerner.com and his Mr Citymen. Been doing som tests and everything seems to be working fine, including putting this guy into the live footage using Matchmover Pro 4.0. Everything will be shot in HD 1920x1080 wich will demand som rendering power. And I have some questions regarding this whole process.
1. We currently dont have any kind of render farms inhouse, therefore we are going to invest in 2-3 computers to set up a small render farm, Ive done some research and hopefully Ive got it right. Im pretty sure that theese computers dont need to be top-notch, therefore buying 2-3 cheap computers for the rendering altogether for under 1000$, all with 3ghz, because we will only be using theese computers for the rendering the spec. dont really seem to matter, or do they?
What are youre suggestions on a simple render farm, is this enough?
2. To set up a render farm, Ive done research on this and it really seems like a pain in the ass to get all things together, but hopefully our solution will work good enough, for the rendering process we will be using Mental Ray with multiple pass rendering, Our solution is to install Maya 8.5 on all computers and setting up the network preferences, simple and easy, but will this work?
3. Compositing, we will be using After Effects to composite it all together, after rendering out our character on the alpha channel we will composite it with the live footage. And also do all the color correction in here. In some of the shots people will be passing by or perhaps a table will stand in the way for our character, to solve this Im thinking about rotoscoping, (frame-by-frame) with mask or vector paint on the matte channel in after effects. Is this totally wrong? should we do this in maya instead? I dont think so, but Im hearing of new techniques every day and just wondering how a professional studio solves this issue. Any suggestions?
If you have any suggestions or comments on our workflow please dont hesitate to post them!
/Daniel Lundh
www.actionfiction.se
We have just been assigned a big job including rigging and building + lipsynch character and putting this into live footage, handheld camera. The spot will be approx 3 minutes long including separate graphics and we will also be in charge of the shoot on location. What we want to achieve is something similar to www.ericlerner.com and his Mr Citymen. Been doing som tests and everything seems to be working fine, including putting this guy into the live footage using Matchmover Pro 4.0. Everything will be shot in HD 1920x1080 wich will demand som rendering power. And I have some questions regarding this whole process.
1. We currently dont have any kind of render farms inhouse, therefore we are going to invest in 2-3 computers to set up a small render farm, Ive done some research and hopefully Ive got it right. Im pretty sure that theese computers dont need to be top-notch, therefore buying 2-3 cheap computers for the rendering altogether for under 1000$, all with 3ghz, because we will only be using theese computers for the rendering the spec. dont really seem to matter, or do they?
What are youre suggestions on a simple render farm, is this enough?
2. To set up a render farm, Ive done research on this and it really seems like a pain in the ass to get all things together, but hopefully our solution will work good enough, for the rendering process we will be using Mental Ray with multiple pass rendering, Our solution is to install Maya 8.5 on all computers and setting up the network preferences, simple and easy, but will this work?
3. Compositing, we will be using After Effects to composite it all together, after rendering out our character on the alpha channel we will composite it with the live footage. And also do all the color correction in here. In some of the shots people will be passing by or perhaps a table will stand in the way for our character, to solve this Im thinking about rotoscoping, (frame-by-frame) with mask or vector paint on the matte channel in after effects. Is this totally wrong? should we do this in maya instead? I dont think so, but Im hearing of new techniques every day and just wondering how a professional studio solves this issue. Any suggestions?
If you have any suggestions or comments on our workflow please dont hesitate to post them!
/Daniel Lundh
www.actionfiction.se
